The present invention relates to metal oxide semiconductor (MOS) image sensors and, more particularly, to reset circuits for active pixel sensors (APS) in a CMOS array.
Integrated circuit technology has revolutionized various fields including computers, control systems, telecommunications, and imaging. There are a number of types of semiconductor imagers, including charge coupled devices, photodiode arrays, charge injection devices, and hybrid focal plane arrays. Some sensors are referred to as active pixel image sensors (APS). An active pixel image sensor is defined as an image sensor technology that has one or more active transistors within the pixel unit cell. Some types of active pixel sensor technologies include the amplified MOS imager (AMI), charge modulation device (CMD), volt charge modulated device (VCMD), base stored image sensor (BASIS), and the static induction transistor (SIT).
One prior art circuit using a CMOS photodiode-type active pixel sensor is shown in xe2x80x9c128xc3x97128 CMOS Photodiode-Type Active Pixel Sensor With On-Chip Timing, Control And Signal Chain Electronics,xe2x80x9d by R. H. Nixon et al., Proceedings of the SPIE-The International Society for Optical Engineering, Volume 2415, 1995, pages 117-123. FIG. 2 of that reference has been reproduced herein as FIG. 1.
FIG. 1 is a schematic diagram of a CMOS APS, along with readout circuits. The pixel unit cell 25 consists of a photodiode (PD), a source-follower input transistor SF1, a row-selection transistor ROW and a reset transistor RESET which controls lateral blooming through proper biasing of its gate. The drain diffusion 50 is coupled to the voltage source VDD. At the bottom of each column of pixels, there is a column circuit 60 consisting of a load transistor VLN1 and two output branches to store the reset and signal levels. Each branch consists of a sample and hold capacitor (CS or CR) with a sampling switch (SHS or SHR) and a second source-follower SF2 with a column-selection switch (COL). The reset and signal levels are read out differentially, allowing correlated double sampling to suppress 1/f noise and fixed pattern noise (not kTC noise) from the pixel. A double delta sampling (DDS) circuit shorts the sampled signals during the readout cycle reducing column fixed pattern noise. These readout circuits 60 are common to an entire column of pixels. A read-out circuit 70 that is common to the entire array includes load transistors (VLN2) of the second set of source followers (VLP) and the subsequent clamp circuits CLAMP and output source followers SF3 and SF4.
Timing for the readout sequence is as follows. After a row has been selected, the signal that is present on each column pixel in that row is sampled (SHS) onto the holding capacitor CS. Next, each pixel in the row is reset (RESET). This is followed by sampling the reset level (SHR) onto holding capacitor CR. A simplified expression for the output voltage of the reset branch of the column circuit is given by:
Vcolxe2x80x94R≅xcex2{xcex1[Vpdrxe2x88x92Vtpix]xe2x88x92Vtcolr}
where xcex1 is the gain of the pixel source-follower, xcex2 is the gain of the column source-follower, Vpdr is the voltage on the photodiode after reset, Vtpix is the threshold voltage of the pixel source-follower n-channel transistor, and Vtcolr is the threshold voltage of the column source-follower p-channel transistor. Similarly, the output voltage of the signal branch of the column circuit is given by:
Vcolxe2x80x94S≅xcex2{xcex1[Vpdsxe2x88x92Vtpix]xe2x88x92Vtcols}
where Vpds is the voltage on the photodiode with the signal charge present and Vtcols is the threshold voltage of the column source-follower p-channel transistor. Experimentally, the peak to peak variation in Vtcolrxe2x88x92Vtcols is typically 10-20 mV. It is desirable to remove this source of column-to-column fixed pattern noise FPN. JPL has previously developed a double delta sampling (DDS) technique to eliminate the column-to-column FPN. This approach represents an improved version of the DDS circuitry.
Sequential readout of each column is as follows. First a column is selected. After a settling time equivalent to one-half the column selection period, the DDS is performed to remove column fixed pattern noise. In this operation, a DDS switch and two column selection switches on either side are used to short the two sample and hold capacitors CS and CR. Prior to the DDS operation the reset and signal column outputs (Vcol_R and Vcol_S) contain their respective signal values plus a source follower voltage threshold component. The DDS switch is activated immediately after CLAMP is turned off. The result is a difference voltage coupled to the output drivers (VR_OUT and VS_OUT) that is free of the voltage threshold component.
One problem of this circuit is that the implementation of the active pixel cell 25 continues to have the problems related to the design choice between hard reset and soft reset implementations. The choice between these implementations determine how the reset signal RST and the voltage on the drain diffusion 50 (hereafter designated as VRST) will be controlled. In the implementation illustrated in FIG. 1, a soft reset implementation is shown, in that the voltage VRST on the drain diffusion 50 is shown to be VDD. This is a typical prior art implementation in which VRST=VDD=RST. One definition of a soft reset system is where the signal (shown as VDD on the drain diffusion 50 in FIG. 1) is greater than the reset voltage RST, minus a threshold voltage VT. This soft reset definition is illustrated in the equation:
soft reset: VRST greater than RSTxe2x88x92VTxe2x80x83xe2x80x83(1)
where the threshold voltage VT is a variable, depending on sensor potential. Also, the theoretical equation for the noise in a soft reset system is:                               soft          ⁢                      xe2x80x83                    ⁢          reset          ⁢                      :                    ⁢                      xe2x80x83                    ⁢          Noise                =                              1            2                    ⁢                                    KT              C                                                          (        2        )            
In contrast, one definition of a hard reset system is where the VRST level is less than the reset signal RST minus the threshold voltage VT. This hard reset definition is illustrated in the equation:
xe2x80x83hard reset: VRST less than RSTxe2x88x92VTxe2x80x83xe2x80x83(3)
And the theoretical equation for the noise in the hard reset method is:                               hard          ⁢                      xe2x80x83                    ⁢          reset          ⁢                      :                    ⁢                      xe2x80x83                    ⁢          Noise                =                              1            2                    ⁢                                    KT              C                                                          (        4        )            
The hard reset and soft reset equations (1) and (3) illustrate that there is a critical voltage VCR which marks the boundary between a soft reset and a hard reset. This critical voltage is defined by the equation:
critical voltage: VCR=RSTxe2x88x92VTxe2x80x83xe2x80x83(5)
The advantages of the hard reset method is that it is fast and uniform, but in theory it suffers from approximately twice the reset noise problems (kTC noise) of the soft reset method. In contrast, the soft reset method has less reset noise, but causes image lag, because the reset is incomplete due to the reset transistor operating in the subthreshold region.
The present invention is directed to a method and apparatus that overcome the foregoing and other disadvantages. More specifically, the present invention is directed to a method and apparatus for reducing image lag through an improved soft reset circuit for an active pixel sensor (APS).
An improved active pixel sensor soft reset circuit for reducing image lag is provided. The active pixel sensor circuit includes a sensor for outputting a sensor potential, and a reset transistor for resetting the sensor. A buffer transistor buffers the output of the sensor, and a row select transistor is used for the read-out function. The row select transistor is coupled between the buffer transistor and a bit line. The bit line is also coupled to a loading transistor.
In accordance with one aspect of the invention, the sensor potential is pulled down to a sufficiently low level during a pull down function that may be implemented before and/or during the soft reset function. If the sensor potential is pulled down during the soft reset function, the pull down time is made to be less than the soft reset time. The low level is set between 0 and the critical potential at which the reset transistor will be on when the soft reset function starts. The timing of the pull down function should stabilize the sensor at the low potential before the actual soft reset function begins.
In accordance with another aspect of the invention, the sensor potential is pulled down by a pull-down circuit. In one embodiment, the pull-down circuit may consist of an NMOS transistor and PMOS transistor, coupled together as a CMOS inverter.
In accordance with another aspect of the invention, the sensor potential may be pulled down by the bit line. The bit line in turn, may be pulled down in a number of different ways. One way for the bit line to be pulled down is through natural discharge. A way to increase the speed with which the bit line is pulled down is to increase the bias on the loading transistor. One way to increase the bias on the loading transistor is to use a special biasing circuit to increase the bias. Another way to increase the bias on the loading transistor is to use a pull-up transistor, which may be either an NMOS transistor or a PMOS transistor.
It will be appreciated that the disclosed method and apparatus for an improved active pixel sensor soft reset circuit is able to obtain the advantages of a hard reset, while maintaining the low reset (kTC) noise of a soft reset. The image lag problem of other soft reset implementation methods is reduced. Furthermore, the active pixel sensor may be implemented using any suitable technology, such as photodiode, photogate, or pinned diode.